.
Angewandte
Communications
We were intrigued by recent reports published by Winne
Presently, we cannot conclusively rule out the possibility that
small amounts of adventitious Brønsted acid, resulting from
hydrolysis of the catalyst, may at least be partially responsible
[
10]
and Pattenden.
They treated 2-furfuryl alcohols with
a stoichiometric amount of TiCl to promote the formation
4
[13]
of furfuryl cations for (4+3) cycloaddition reactions. We
surmised that nucleophilic addition of the C3 of an indole to
either an aldehyde or ketone would furnish the alcohol 9
for promoting the reactions in the entries using Lewis acids.
However, several reports by Olah, Prakash, and co-workers
[12]
describing the hydrolytic stability of gallium(III) salts lead
us to consider such a scenario to be unlikely.
(
Scheme 1). In the presence of an appropriate Lewis acid,
water would be ejected to generate the indolyl cation 10 for
The scope of the three-component (4+3) reaction appears
to be quite broad (Scheme 3). We surveyed a variety of
combinations involving the indoles 5a–h, aldehydes 6a–e and
16, ketal 15, ketones 17a–c, and dienes 7a–e. The reaction is
tolerant to NÀH indoles as well as indoles having an alkyl
[10]
(
4+3) reactions analogous to those reported by Winne et al.
Our group has recently been interested in developing
[
11]
gallium(III)-catalyzed processes.
Because gallium(III)
are stable to air and
[
12]
salts, and in particular Ga(OTf)3,
moisture, we hypothesized that they could be used in
a catalytic fashion despite the fact that an equivalent of
water would be generated in the proposed three-component
group on the nitrogen atom. Electron-withdrawing or elec-
tron-donating groups on either the indole or carbonyl
components are compatible. Electron-rich substrates such as
5d, 6b, and 6d,e appear to accelerate the reaction rate. With
(
4+3) cycloaddition.
As is evident from Table 1, both Ga(OTf) and GaBr3
less reactive substrates the use of Ga(OTf) was beneficial for
3
3
were effective in promoting the desired reaction. Ga(OTf)3
was qualitatively the more reactive catalyst. But because it
achieving higher yields and shorter reaction times (8i, 8k–n).
The presence of halides in 8s–u, offers a convenient handle
for additional elaboration of the indole core through tran-
sition metal-catalyzed cross-coupling reactions.
[
a]
Table 1: Optimization studies.
For the products 8a–f, diastereoselectivities were
observed in the range from greater than 10:1 to 1:1.
Electron-rich aldehydes appear to provide products with the
highest selectivities while electron-poor and linear alkyl
substrates exhibited lower levels of selectivity. The use of an
unsymmetrical diene, isoprene (7c), furnished the desired
product 8m as a single regioisomer. The structures of 8j, 8n,
and 8p were confirmed by single-crystal X-ray analysis.
Unfortunately, neither dimethoxymethane nor formalde-
hyde were suitable substrates. Attempts to extend the diene
scope to include pyrrole (and several N-protected deriva-
tives), furan, thiophene, Danishefskyꢀs diene, or Rawal–
Kozminꢀs diene were not successful. What we observed in
most cases was the addition of the diene (i.e. formation of
a single CÀC bond) to the indolyl cation rather than the
[
e]
Entry
Catalyst
Yield [%]
Comments
[
b]
1
2
3
4
5
6
7
8
9
1
1
1
1
Ga(OTf)3
GaBr3
GaBr3
GaBr3
GaCl3
91
94
84
–
by-product hard to separate
[
[
c]
d]
only 9 observed
only 5a and 9 observed
substantial decomposition
–
In(OTf)3
InBr3
54
92
32
14
–
–
52
83
InCl3
mostly 9 observed
slow reaction
only 9 observed
only 9 observed
slow reaction
Sc(OTf)3
Cu(OTf)2
TFA
TsOH
TfOH
0
1
2
3
desired cycloaddition products.
To gain insight into the mechanism of this reaction, DFT
[
b]
[14]
by-product hard to separate
calculations were performed using the M06 functional and
[
15]
the 6-311G** ++ basis set, as implemented in the Jaguar
[
a] 10 mol% catalyst, 2 equiv 6a, 5 equiv 7a, RT, CH Cl (all rxns stopped
at 2 h). [b] Reaction time 35 min. [c] Toluene as solvent. [d] THF or Et O
as solvent. [e] Diastereoselectivities were 3:1 to 5:1 for all reactions in
which product was observed (as determined by H NMR spectroscopy).
2 2
[16]
program. For the reaction of cyclopentadiene (7a) with the
indolyl cation derived from indole (5c) and the ketal 15, the
lowest energy pathway was identified as a two-step process.
Rather than a concerted pericyclic reaction, an intermediate
is formed by initial CÀC bond formation between the terminal
2
1
TFA=trifuoroacetic acid, Tf=trifluoromethanesufonyl, Ts=4-toluene-
sulfonyl.
carbon atom of the indolyl cation and a terminal diene carbon
atom, with subsequent closure to the seven-membered ring.
Transition states for both steps were located, and an overall
reaction free-energy profile is depicted in Figure 1 (full details
are provided in the Supporting Information). A stepwise
mechanism is in agreement with what Winne et al. proposed
also generated a by-product (not formed with the use of
GaBr ) which was difficult to separate from 8a, we decided to
3
explore the scope of the transformation using GaBr . InBr3
3
was effective as well but several other Lewis acids (entries 6
and 8–10) resulted in either diminished yields or varying
amounts of the alcohol 9.
[
10a]
for the related (4+3) reaction of furfuryl cations.
We also
We also examined the use of three Brønsted acids. TFA
Table 1, entry 11) generated only 9 but TsOH and TfOH
entries 12 and 13) furnished the desired products in 52 and
3% yields, respectively. While the yield with TfOH
approaches that obtained with gallium(III) salts for this
particular combination of substrates, TfOH is markedly less
effective with less reactive dienes (see 8l–n in Scheme 3).
explored an alternative pathway in which an initial (4+2)
cycloaddition might be followed by ring expansion, by 1,2-
migration, to the observed product. No low-energy synchro-
nous (4+2) pathway was located; the formation of the
intermediate in Figure 1 is the lowest energy initial step.
Closure of the same intermediate to give the formal (4+2)
product proceeds via a transition state that is higher in energy
(
(
8
2
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!